Alcohol consumption is a top-10 contributor to the worldwide burden of disease. In 2012, the Monograph program at the International Agency for Research on Cancer concluded that cancers of the upper aero-digestive tract (UADT; oral cavity, pharynx, larynx, esophagus), liver, colorectal and female breast were causally related to alcohol consumption. The evidence of a link with other cancers, primarily prostate and pancreatic cancer, but also kidney and thyroid cancer, melanoma, and Hodgkin (HL) and non-Hodgkin (NHL) lymphomas was considered inconsistent or sparse, especially for cohort studies. The following priorities were identified: 1) to assess the role of alcohol in a wider list of cance sites; 2) to elucidate the role of drinking patterns (in particular binge drinking, an increasing factor in many countries) and specific alcoholic drinks on cancer risk; and 3) to elucidate the rol of smoking in alcohol-related carcinogenesis. We address all three priorities in this proposal. We will conduct the largest, most comprehensive pooled analysis on alcohol and risk of specific cancers. These analyses will be conducted in the international Pooling Project of Prospective Studies of Diet and Cancer (DCPP), with >2.6 million women and men in 35 cohorts. We propose to examine alcohol use at enrollment, lifetime alcohol use, age at drinking initiation, and patterns of alcohol intake (including binge drinking) and risk of cancers for which a positive association is suggested (e.g., prostate and pancreatic cancer and melanoma), and risk of cancers for which an inverse association is suggested (e.g., kidney and thyroid cancer, HL, and NHL). Collectively, these studies include >161,000 incident cases of these cancers (expected case numbers: 2,288 for HL; 4,882 for thyroid cancer; 9,940 for kidney cancer, and >10,000 each for pancreatic, UADT and prostate cancer, as well as NHL and melanoma). We hypothesize that associations for lifetime alcohol consumption will be stronger than those for a single measure of alcohol consumption at enrollment and that specific patterns of alcohol drinking (such as excessive/binge drinking) and type of alcoholic beverages will be important determinants of these relationships. Because residual confounding by smoking is a concern in analyses of alcohol intake, we also will conduct analyses among never and past smokers separately (including for UADT cancer). This proposal builds efficiently on the extensive, existing harmonized data on baseline alcohol intake and cancer risk factor data for 29 of the 35 cohorts. Solid new prospective evidence will be produced on the role of alcohol at cancer sites for which the relationship with alcohol is not established, including cancers whose etiology is still poorly understood (e.g., prostate, kidney and pancreatic cancer). As current cancer prevention guidelines recommend limiting alcohol intake to 2 drinks/d for men and 1 drink/d for women, our examination of the role of alcohol at these levels will have an important public health impact, as risk increases associated with light to moderate intakes will potentially affect a large proportion of population.